Serge Vasylechko1,2, Simon K. Warfield1,2, Sila Kurugol1,2, and Onur Afacan1,2
1Boston Children's Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States
1Boston Children's Hospital, Boston, MA, United States, 2Harvard Medical School, Boston, MA, United States
We generated substantial amount of 3D synthetic T2 relaxometry data with a realistic forward model, and demonstrate its application to myelin water fraction. Our network has resulted in an excellent accuracy in the synthetic test dataset, and generated similar MWF maps as the NNLS algorithm.
Figure 2: An example of the generated synthetic data. Top row shows the model parameters, second row shows the generated signals and third row shows generated spatial transformations.
Figure 1: A flowchart detailing the proposed pipeline for generation of large scale 3D synthetic datasets for multi-component T2 distributions within the naturally occuring bounds, with a spatially varying sampling model.